Mbia Insurance Corporation Wells Fargo Bank Minnesota, N.A., as Trustee of Sfc Grantor Trust, Series 2000-1, Sfc Grantor Trust, Series

OPINION OF THE COURT ALITO, Circuit Judge. In these consolidated appeals, we are called upon to construe a series of contracts. Appellant Royal Indemnity Company (“Royal”) agreed in these contracts to insure the repayment of principal and interest on several hundred million dollars of student loans. The named beneficiaries of the policies, Wells Fargo Bank Minnesota, N.A. (‘Wells Fargo”) and Wilmington Trust of Pennsylvania (Wilmington Trust”), 1 sued after the loans went into default and Royal failed to pay the claims they submitted. Royal defended on the ground that the lender on the underlying obligations fraudulently induced it to issue the policies and that this fraud entitled it to rescission. The District Court entered summary judgment for the beneficiaries, and this appeal followed. We agree with the District Court that Royal’s policies unambiguously and effectively waive defenses to its obligations based on fraud, but we conclude that Royal has raised a triable issue as to whether all of the losses claimed by the beneficiaries were covered under its *208 policies. We thus affirm in part and reverse in part. I. Student Finance Corporation (“SFC”) was founded in 1992 to cater to the vocational segment of the student loan market. Some of the loans it originated itself; others it acquired from the original lenders. Many of the loans were apparently made to students at truck-driving schools. At all relevant times, SFC was owned and controlled by its founder and chief executive officer, Andrew N. Yao. The capital for SFC’s business came from financial institutions like Wilmington Trust and Wells Fargo. In 1999, Wilmington Trust issued SFC a $75 million loan, taking a pool of student loans as security. Wells Fargo, by contrast, helped finance SFC by securitizing the older loans in its portfolio. In each securitization, student loans were packaged and sold to a trust settled for the specific purpose of holding title to the loans. Wells Fargo, as trustee, funded the purchase by selling certificated shares in the trust to institutional investors. The record reflects that Wells Fargo raised approximately $450 million for SFC in eight securitizations from 1999 to 2002. To encourage Wilmington Trust and the investors in Wells Fargo’s trusts to part with their capital, SFC contracted with Royal to insure the repayment of interest and principal on the student loans. At issue in this case are ten “Credit Risk Insurance Policies” issued by Royal. Eight of them insured the loans held by the Wells Fargo trusts (one per trust), and two of them insured the loans pledged as collateral to Wilmington Trust. Each policy named either Wells Fargo or Wilmington Trust as beneficiary. The following table summarizes their terms: Table 1. Summary of Policy Terms Policy Number_Beneficiary Inception Liability Limit RST 293334 Wells Fargo 1/22/99 $ 50,000,000.00 RST 293309 Wells Fargo 12/3/99 $ 53,053,642.08 RST 147522 Wells Fargo 4/30/00 $ 48,459,255.76 RST 147524 Wells Fargo 8/30/00 $ 29,999,999.94 RST 147525 Wells Fargo 11/27/00 $ 55,616,550.00 RST 147526 Wells Fargo 1/31/01 $ 48,286,713.44 RST 147538 Wells Fargo 10/19/01 $120,000,000,00 RST 147536 Wells Fargo 11/15/01 $ 80,000,000.00 Wilmington RST 321276 Trust 1/22/99 $ 76,000,000.00 Wilmington RST 147533 Trust_8/17/01 $ 5,518.459.00 Royal alleges, and the beneficiaries do not dispute, that SFC procured this insurance through a spectacular fraud. According to Royal, SFC misrepresented the creditworthiness and employment history of its student borrowers and conspired with schools to generate as many loans as possible by altering or forging loan documents. As loans went into default, SFC allegedly paid some of them down by surreptitiously diverting the proceeds of later loans. By thus masking the default rates of the older loans, SFC allegedly induced Royal to insure still more loans, whose proceeds then had to be applied in Ponziesque fashion to pay down the earlier ones. According to Royal, some of the proceeds were also diverted to Yao’s personal accounts. SFC’s business proved unsustainable. In a March 2002 telephone call to Royal, Yao allegedly confessed that SFC had been paying down defaulted loans and explained that this practice could no longer be continued. An investigation launched by Royal revealed that SFC’s loans had been defaulting at rates — over 80% in the case of one pool — well in excess of reported figures. Within three months of Yao’s call, SFC was in Chapter 7 bankruptcy, where it apparently remains to this day. With SFC no longer paying down student loans, the defaults fell on the shoulders of Wilmington Trust and Wells Fargo, who turned to Royal to make good on its poli *209 cies. Claims on those policies, which totaled only $38.6 million between 1999 and the spring of 2002, had piled up to $380 million by the end of 2002. A flurry of lawsuits followed. Royal filed suit in Texas state court to rescind the policies, but this ease was dismissed after limited discovery for lack of personal jurisdiction. In July 2002, Wells Fargo and MBIA Insurance Corporation (“MBIA”) 2 sued Royal in the U.S. District Court for the District of Delaware. Their complaint, which invoked the Court’s diversity jurisdiction, alleged that Royal had wrongfully repudiated the trusts’ eight policies by filing the Texas action and sought relief in the form of specific performance, a declaratory judgment that the policies were in effect, and damages. Wilmington Trust filed its own suit in Delaware federal court, alleging essentially the same claims. Royal defended on the ground that SFC’s fraudulent inducement entitled it to rescission. The beneficiaries countered that the policies unambiguously waived this defense. In separate opinions, the District Court entered summary judgment for the beneficiaries. See MBIA Ins. Corp. v. Royal Indem. Co., 312 F.Supp.2d 583 (D.Del.2004); MBIA Ins. Corp. v. Royal Indem. Co., 286 F.Supp.2d 347 (D.Del.2003). It found that the ten policies unambiguously waived fraud in the inducement as a defense to payment, and it predicted that Delaware’s highest court would enforce them. The Court acknowledged that Delaware law was reluctant to enforce boilerplate waivers against unsophisticated parties, but it believed a clear waiver negotiated by sophisticated parties could be enforced. See 312 F.Supp.2d at 586, 286 F.Supp.2d at 355. Since the Court concluded that Royal’s waivers were enforceable, and that they clearly covered the defense Royal sought to present, it found further discovery on that defense unnecessary. It awarded summary judgment to the beneficiaries, ordering Royal to pay $269,851,527 plus interest to Wells Fargo and $12,908,966.43 plus interest to Wilmington Trust, and it further ordered Royal to pay subsequent claims as they came due. Royal timely appealed to this Court. II. We review an award of summary judgment de novo, applying the same test on review that the District Court should have applied. In re Ikon Office Solutions, Inc., 277 F.3d 658, 665 (3d Cir.2002). Summary judgment should be awarded when “the pleadings, depositions, answers to interrogatories, and admissions on file, together with the affidavits, if any, show that there is no genuine issue as to any material fact and that the moving party is entitled to a judgment as a matter of law.” Fed.R.Civ.P. 56(c). All reasonable inferences from the record must be drawn in favor of the nonmoving party. Brewer v. Quaker State Oil Refining Corp., 72 F.3d 326, 333 (3d Cir.1995). The Court may not weigh the evidence or assess credibility. Boyle v. County of Allegheny Pa., 139 F.3d 386, 393 (3d Cir.1998). Royal attacks the District Court’s award of summary judgment on several grounds. It first argues that the Court erred in finding that the text of its policies unam *210 biguously waived its defense to payment based on SFC’s fraud. It further argues that waivers of the breadth and generality found in its policies are unenforceable under Delaware law. Finally, it argues that triable issues remain as to whether the policies covered all of the losses claimed by the beneficiaries. Since it is undisputed that Delaware provides the substantive law for this dispute, we turn to that state’s law of contract to determine whether summary judgment was properly awarded. III. A. Delaware follows the objective theory of contract. See Haft v. Haft, 671 A.2d 413, 417 (Del.Ch.1995); Progressive Int’l Corp. v. E.I. Du Pont de Nemours & Co., No. C.A. 19209, 2002 WL 1558382, at *7 (Del.Ch.2002) (unpublished opinion). Although the law of contract generally strives to enforce agreements in accord with their makers’ intent, the objective theory considers “objective acts (words, acts and context)” the best evidence of that intent. Haft, 671 A.2d at 417. Unambiguous written agreements should be enforced according to their terms, without using extrinsic evidence “to interpret the intent of the parties, to vary the terms of the contract or to create an ambiguity.” Eagle Indus., Inc. v. DeVilbiss Health Care, Inc., 702 A.2d 1228, 1232 (Del.1997); see also City Investing Co. Liquidating Trust v. Cont’l Cas. Co., 624 A.2d 1191, 1198 (Del.1993). A contract is not ambiguous merely because the parties disagree about its proper interpretation. Whether a contract is ambiguous is determined according to an objective, reasonable-person standard and is a question of law. See Eagle Indus., 702 A.2d at 1233 n. 8 (“The true test is what a reasonable person in the position of the parties would have thought it meant.”). Words are to be given their ordinary meaning and should not be “torture[d]” to impart ambiguity where none exists. Rhone-Poulenc Basic Chems. Co. v. Am. Motorists Ins. Co., 616 A.2d 1192, 1196 (Del.1992). The language of Royal’s waivers varies from policy to policy, but we agree with the District Court that each policy satisfies these standards of clarity. We consider the contracts in reverse chronological order, beginning with the last two policies Royal issued and considering the remaining eight in subsection 2 below. 1. Policies RST 147536 and RST 147538 are the clearest. Policy RST 147536 states in bold capital letters: THE RIGHT OF THE BENEFICIARY TO RECEIVE PAYMENT FOR LOSSES UNDER THIS POLICY SHALL BE ABSOLUTE, CONTINUING, IRREVOCABLE AND UNCONDITIONAL IRRESPECTIVE OF (A) ANY FRAUD WITH RESPECT TO THE STUDENT LOANS, (B) THE GENUINENESS, VALIDITY OR ENFORCEABILITY OF ANY ... STUDENT LOAN OR THE BREACH OF ANY SUCH CONTRACT OR ANY COVENANT OR REPRESENTATION OR WARRANTY MADE THEREIN, OR (C) ANY OTHER RIGHTS OR DEFENSES THAT MAY BE AVAILABLE TO THE INSURER TO AVOID PAYMENT OF ITS OBLIGATION UNDER THIS POLICY (ALL OF WHICH RIGHTS AND DEFENSES ARE HEREBY EXPRESSLY WAIVED BY THE INSURER).... App. at 2221. The language of Policy RST 147538 is virtually identical. See App. at *211 2253-54. 3 The policies plainly strip Royal of a defense to payment based on fraud, first by declaring that its liability will be unaffected by “fraud with respect to the student loans” and then by “expressly waiv[ing]” any other right or defense it could marshal to avoid payment. In the face of this clarion language, Royal contends that the phrase “with respect to the student loans” contemplates only the microfraud of individual schools or students, not the macrofraud of SFC. The waivers were designed, argues Royal, only to protect Wells Fargo from obstreperous litigation over the validity of individual loan documents. In support of this position, Royal points to clause (B), which homes in on “the genuineness, validity or enforceability of any ... student loan.” Invoking the expressio unius and ejusdem generis canons, it argues that clause (B) waives only a defense based on the “validity or enforceability” of a student loan and that clause (A) should be given a commensurately narrow meaning. Aside from completely ignoring the blanket language of clause (C) (clause (D) in Policy RST 147538), Royal’s argument renders clause (A) surplusage. If clause (A) were limited to individual cases of fraud by schools or students, it would sweep no further than clause (B). The fraudulent inducement of a loan agreement by a student or school would render the defrauded party’s obligations under that agreement voidable. See Restatement (Second) of Contracts § 164(1) (1981). In such a case, the “genuineness, validity or enforceability” of the loan would be affected. While the validity of the loans would not be affected by SFC’s (or the schools’) fraud on Royal, this is precisely the type of fraud that Royal now claims is not covered by clause (A), despite the fact that it clearly occurs “with respect to” the student loans within the ordinary meaning of those words. See Webster’s Third New International Dictionary 1934 (Philip Babcock Gove ed. in chief, unabr. ed.1971). Since Royal’s interpretation makes clause (A) surplusage and reads out clause (C) (clause (D) in Policy RST 147538), it must be rejected. Royal argues that the definition of “Student Loans” creates a triable issue as to whether the defense of fraudulent inducement was intended to be waived in a case such as this. The definition of “Student Loans” is restricted by each policy to the loans covered therein and does not include any earlier loans SFC may have issued. Since the definition is so restricted, clause (A) could not be read, Royal argues, as waiving defenses based on misrepresentations about the earlier loans on which Royal relied. Even if this policy wrinkle created a triable issue as to the scope of clause (A), it would not follow that summary judgment was unwarranted. To establish fraudulent inducement, Royal must show reasonable and detrimental reliance on a misrepresentation intentionally or recklessly made to induce action or inaction. See Kronenberg v. Katz, 872 A.2d 568, 585 n. 25 (Del.Ch.2004); Gloucester Holding Corp. v. U.S. Tape & Sticky Prods., LLC, 832 A.2d 116, 124 (Del.Ch.2003). Even if clause (A) did not waive a defense based on' misrepresentations with respect to earlier loans, Royal could not establish reliance on those representations in light of the clear “anti-reliance” language of clause (C) (clause (D) in Policy RST 147538). See *212 Kronenberg, 872 A.2d at 593. It is unfathomable that an insurer that intended to rely on extracontractual representations would agree that its obligations are “absolute, continuing, irrevocable and unconditional irrespective of ... any other rights or defenses that may be available to the insurer ... (all of which rights and defenses are hereby expressly waived by the insurer).” App. at 2221, 2254 (emphasis omitted). Royal cannot possibly claim that its reliance on those representations was reasonable when it waived all defenses based on reasonable reliance. Since Royal’s reliance was unreasonable however “Student Loans” is defined, the definition of that term does not create an ambiguity in the policies, and the District Court did not err in finding Royal’s waiver unambiguous. 2. Each of the remaining eight policies contains some variant of the following language: NOTWITHSTANDING ANY OTHER PROVISION OF THIS POLICY TO THE CONTRARY, THE RIGHT OF THE BENEFICIARY TO RECEIVE PAYMENT FOR LOSS UNDER THIS POLICY AFTER PAYMENT OF THE INITIAL PREMIUM BY THE INSURED SHALL BE ABSOLUTE AND UNCONDITIONAL, AND NO FAILURE ON THE PART OF THE INSURED OR THE BENEFICIARY TO OBSERVE OR PERFORM ANY COVENANT OR CONDITION CONTAINED IN THIS POLICY (INCLUDING WITHOUT LIMITATION THOSE CONTAINED IN ARTICLE III, ARTICLE IV AND ARTICLE V) SHALL ENTITLE THE INSURER TO ANY RIGHT OF SET-OFF, COUNTERCLAIM OR DEFENSE AGAINST THE BENEFICIARY OR ANY OTHER PARTIES OR OTHERWISE RELIEVE THE INSURER OF ANY LIABILITY TO MAKE ANY SUCH PAYMENT FOR LOSS TO THE BENEFICIARY UNDER THIS POLICY, SUBJECT ONLY TO THE LIMIT OF LIABILITY. E.g., App. at 5485 (original emphasis). In addition, seven of the policies contain some variant of the following language: The Insurer’s obligation to pay any Claim made under this Policy is absolute, unconditional and irrevocable and shall not in anyway be affected, mitigated or eliminated by (y) a breach of any representation or warranty made by the Insured, the Servicer, Student Finance Corporation or the Beneficiary, or (z) the failure of the Insured or Student Finance Corporation to comply with the Underwriting Policies. E.g., App. at 2021. The eighth policy, RST 321276, refers only to “the Insured,” omitting mention of “the Servicer,” “Student Finance Corporation,” and “the Beneficiary.” App. at 5499-500. Though none of the eight policies uses the word “fraud,” no reasonable interpretation can be teased from this language that would preserve Royal’s defense. The policies all declare that, after payment of the premium, Royal’s liability will become “absolute” and “unconditional” and “subject only to the limit of liability.” Though such language might not contemplate a fraud that induced the policies in the first place, the policies also provide that Royal’s liability will be unaffected by “a breach of any representation.” Since a misrepresentation is an essential element of a fraud, it follows that fraud cannot affect Royal’s liability. Finally, the policies provide that Royal’s liability is unaffected by a “failure ... to comply with the Underwriting Policies.” This is a significant recital because the alleged fraud had its genesis in misrep *213 resentations that SFC’s lending complied with certain underwriting policies. Royal cannot claim that it was entitled to rely on SFC’s representations about its lending when it agreed to liability regardless of whether the lending conformed to those representations. Royal argues that a “breach of ... representation” is not the same thing as a “misrepresentation” but refers instead to the failure of a condition set forth “in the contract or perhaps in specific accompanying contracts.” Royal’s Br., No. 03-4382, at 41. Although the policies do not list any “representations and warranties” on which the parties may rely, Royal observes that the term sheet for each policy declares that “the statements in this Policy ... are the agreements and representations of the Insured” and that “this Policy embodies all agreements existing between the Insured and the Insurer or any of its representatives.” E.g., App. at 5525. According to Royal, the phrase “breach of any representation” must refer to representations in the policies, since those are the only “agreements and representations of the Insured.” It is ironic that Royal should look to an integration clause for evidence that it was entitled to rely on SFC’s extraeontraetual representations. Such an interpretation stands that clause on its head. Even under Royal’s contorted reading, however, the clause makes no mention of SFC’s representations. SFC and the Servicer were not even parties to the policies, so their representations would have to be found outside the four corners of the agreements. In short, the phrase “breach of any representation ... [by] Student Finance Corporation” must refer to extra-contractual representations. Under the only reasonable interpretation of the policies, Royal has agreed that its liability will be unaffected by reliance on SFC’s extra-contractual representations. Royal argues that even if the analysis above is correct it has not waived its defense of fraudulent inducement in Policy RST 321276 because that policy refers only to a breach of representation by “the Insured.” The Insured on that policy was “SFC Financial I, LLC,” a special purpose entity affiliated with SFC.App. at 5477. Royal argues that since it relied on the misrepresentations of SFC rather than that special purpose entity, its common law defense of fraud in the inducement is unaffected by the policy. As we noted earlier, an agreement may foreclose a fraud defense not only by waiving “fraud” but also by setting forth terms clearly inconsistent with reasonable reliance on extraeontraetual representations. Royal here has effectively disclaimed reliance on SFC’s representations no less than those of SFC Financial I, LLC, even though the former is not mentioned in the agreement. Royal admits that both entities were under the ownership and control of Andrew Yao. See, e.g., Royal’s Memorandum in Support of Its Motion for an Order Appointing a Chapter 11 Trustee at 5, App. at 4395 (“Yao ... owned and controlled [SFC] and all of its affiliates at all times pertinent to this case.”). Since SFC Financial I, LLC, was an alter ego of SFC separated only by corporate formalities, Royal’s disclaimer of reliance on the former’s representations made it unreasonable as a matter of law for it to rely on the representations of the latter. Finally, Royal argues that even if the policies waive defenses based on fraud, they do not waive a defense based on the invalidity of the policies themselves. This argument does not identify a reasonable alternative construction of the policies so much as an alternative characterization of Royal’s defense. Royal supposes that the *214 policies are invalid because it was fraudulently induced to issue them. However recharacterized, this remains a defense based on fraud and is waived no less than a defense that invoked SFC’s “fraud with respect to the student loans” or “a breach of [its] representation.” 3. Because we conclude that the agreements on their face waive Royal’s fraud defense in unambiguous terms, we need not consider the extrinsic evidence submitted by the parties. 4 We also need not discuss the doctrine of contra profer-entem, which operates against the insurer only when policy terms are ambiguous. See Twin City Fire Ins. Co. v. Del. Racing Ass’n, 840 A.2d 624, 630 (Del.2003). Finally, we need not decide whether the parties’ agreements are properly characterized as insurance policies or guaranties, since the interpretive principles discussed above apply to both and would yield the same result in either case. See Westfield Ins. Group v. J.P.’s Wharf, Ltd., 859 A.2d 74, 76 (Del.2004); Universal Studios Inc. v. Viacom Inc., 705 A.2d 579, 589 (Del.Ch.1997); see also Restatement (Third) of the Law of Suretyship and Guaranty §§ 1-2 (1996) (explaining that a finding of surety-ship status depends on the rights and obligations set forth in the agreement, not the other way around). However characterized, the policies on their face waive Royal’s defense to payment of its obligations based on fraud, and the District Court did not err in so finding. B. We next consider whether these waivers are enforceable. The Delaware Supreme Court has not addressed the standards for effective waiver of a defense based on fraud in the inducement, so we must predict how it would decide this question. See Koppers Co. v. Aetna Cas. & Sur. Co., 98 F.3d 1440, 1445 (3d Cir.1996). The District Court recognized that some authority had looked askance at broad waivers in adhesive form agreements but predicted that the Delaware Supreme Court would enforce an unambiguous waiver negotiated by sophisticated parties. We agree. The Delaware Supreme Court’s closest precedent on this question is its decision in Norton v. Poplos, 443 A.2d 1 (Del.1982). In that case, the plaintiff sought to rescind a real estate transaction by alleging that the seller had negligently misrepresented *215 the land’s zoning. The Court declined to find the claim barred by a standard integration clause in the purchase agreement reciting that the purchaser had not relied on extracontractual representations. Id. at 3. The Court noted that Delaware case law generally did not consider an integration clause a reliable indicium of intent to waive a fraudulent inducement defense. See id. at 6. It concluded, after a survey of cases from other jurisdictions, that the better authorities extended this rule to negligent misrepresentations, allowing the claim to proceed despite the integration clause. See id.. The boilerplate nature of the parties’ agreement weighed heavily in the Norton Court’s calculus. See id. at 7 (“We see no reason why a court of equity should enforce a standard ‘boiler plate’ provision that would permit one who makes a material misrepresentation to retain the benefit resulting from that misrepresentation at the expense of an innocent party.”). Since the scope of Royal’s obligations turns not on a boilerplate merger clause but on waivers sculpted by parties of exquisite legal and financial sophistication, we do not believe Norton provides a reliable guidepost for our decision. See Great Lakes Chem. Corp. v. Pharmacia Corp., 788 A.2d 544, 555 (Del.Ch.2001) (distinguishing the agreements in Norton and its progeny as “simple real estate contracts having boilerplate, unnegotiated disclaimer language”). More reliable guidance may be found in the Chancery Court’s recent decision in Kronenberg v. Katz, 872, A.2d 568 (Del.Ch.2004). In a thoughtful opinion by Vice Chancellor Strine, the Court reasoned that an enforceable waiver of fraud should have “language that, when read together, can be said to add up to a clear anti-reliance clause by which the plaintiff has contractually promised that it did not rely upon statements outside the contract’s four corners in deciding to sign the contract.” Kronenberg, 872 A.2d at 593. The Court acknowledged that an integration clause could be read as a promise of nonreliance, since the clause recited that the contract was the “entire agreement” and superseded all prior “understandings” and “inducements.” See id. at 591. But it found the integration clause also susceptible to a “more traditional interpretation,” according to which the clause “simply operates to police the variance of the agreement by parol evidence.” Id. at 591-92. Since these competing interpretations of the clause left the parties’ intent in doubt, the Court declined to enforce it to bar a fraud claim. Royal protests that the sine qua non of enforceability is not clarity but specificity. A waiver, argues Royal, is nothing but the voluntary relinquishment of a known right. See Kallop v. McAllister, 678 A.2d 526, 532 (Del.1996). Since a waiver must be voluntary, clear evidence of the party’s intent to waive is required for enforcement. See Realty Growth Investors v. Council of Unit Owners, 453 A.2d 450, 456 (Del.1982); George v. Frank A. Robino, Inc., 334 A.2d 223, 224 (Del.1975). A requirement of specificity serves this end by ensuring that parties do not, by promising the moon, unintentionally waive rights they had not really contemplated. Accordingly, Royal concludes, Delaware courts will refuse to enforce even a clear waiver of fraud unless it sets forth the particular representations on which the induced party agreed not to rely. Though this argument has some force, its conclusion finds virtually no support in Delaware law. Cases have focused on the function of the terms at issue, see DRR, L.L.C. v. Sears, Roebuck & Co., 949 F.Supp. 1132, 1137 (D.Del.1996) (holding that an “as is” clause does not “insulate a *216 seller from suit for its fraudulent misrepresentations”); Traylor Eng’g & Mfg. Co. v. Nat’l Container Corp., 70 A.2d 9, 13 (Del.1949) (holding that a warranty cannot serve to disclaim reliance on extracontractual representations for purposes of a fraud claim); DCV Holdings, Inc. v. Conagra, Inc., No. 02-550, 2003 WL 2008199, at *3 (Del. Apr. 29, 2003) (unpublished table decision) (finding an all-inclusive warranty ambiguous with respect to waiver of fraud claims); Kronenberg, 872 A.2d at 592 (“[M]any learned authorities state that typical integration clauses do not operate to bar fraud claims.”), or the circumstances of their negotiation. See Norton, 443 A.2d at 7 (discounting ‘“boiler plate’ found in the merger clause”); Great Lakes, 788 A.2d at 555 (observing the “carefully negotiated disclaimer language” crafted by “highly sophisticated parties, assisted by industry consultants and experienced legal counsel”). Although specificity and clarity often go hand in hand, we are unaware of any Delaware case that has made specificity a prerequisite to effective waiver. In fact, a recent unpublished opinion of the Chancery Court appears to have rejected the specificity requirement. See Progressive Int’l Corp. v. E.I. Du Pont de Nemours & Co., No. C.A. 19209, 2002 WL 1558382 (Del.Ch. July 9, 2002). The Court there held that an integration clause reciting that neither party would rely on extra-contractual promises or representations foreclosed rescission based on fraud. The allegedly defrauded party protested that the clause did not identify the unreliable representations or promises, but the Court found this omission insignificant: If adopted as law, Progressive’s argument would impede commerce. It is not efficient for negotiators to identify all the material issues that are not part of the foundation of their relationship, and to list them in a contractual schedule. Indeed, that exercise would be wasteful and silly, as the integration clause in the License Agreement shows. By simple and unambiguous means, the parties to that Agreement identified all the representations and statements that could not have induced the execution of the Agreement: all representations and statements not included in the text of the Agreement itself. Id. at *10; see also DeBakey Corp. v. Raytheon Serv. Co., No. 14947, 2000 WL 1273317, at *27 (Del.Ch. Aug.25, 2000) (rejecting a fraud counterclaim because the aggrieved party could have included the assurances on which it relied as representations and warranties in the agreement). Although the Progressive decision was not published, we believe its persuasive reasoning provides a good model for the Delaware Supreme Court’s own decision in this area. The putative requirement of specificity seems rooted not in Delaware case law but in a line of New York cases, and even there it has been applied inconsistently. Although the Second Circuit in Manufacturers Hanover Trust Co. v. Yanakas, 7 F.3d 310, 316 (2d Cir.1993), declared specificity the “touchstone” of enforceability, other authority has focused on the clarity of the waiver rather than its detail. See Valley Nat’l Bank v. Greenwich Ins. Co., 254 F.Supp.2d 448 (S.D.N.Y.2003); Citibank, N.A. v. Plapinger, 66 N.Y.2d 90, 495 N.Y.S.2d 309, 485 N.E.2d 974 (1985). The Plapinger Court found the omission of detail unimportant, deeming it “unrealistic in such circumstances to expect an express stipulation that defendants were not relying on a separate oral agreement.” 485 N.E.2d at 977. The Court in Valley National Bank reached a similar conclusion, distinguishing Yanakas as “less applicable ... where the drafter and more sophisticated party in the transaction now claims *217 that the disclaimer is too broad.” 254 F.Supp.2d at 458. Royal relies heavily on JPMorgan Chase Bank v. Liberty Mutual Insurance Co., 189 F.Supp.2d 24 (S.D.N.Y.2002), a case brought over a natural gas forward contract between Enron and one of its affiliates. The affiliate borrowed money from plaintiff bank to purchase the gas. Defendant surety guarantied that Enron would deliver it. Unbeknownst to the surety, the bank had agreed with Enron that the affiliate would subsequently repurchase the gas at a higher price. This secret side agreement transformed the forward contract into a simple loan. See id. at 26. The Southern District of New York concluded that the surety could avoid payment on the bond, rejecting the argument that “a general sweeping disclaimer can serve to disclaim any and all extrinsic fraud between sophisticated parties.” Id. at 27. It read Plapinger and Yanakas to require “a clear indication that the disclaiming party has knowingly disclaimed reliance on the specific representations that form the basis of the fraud claim.” Id. Were JPMorgan Chase a case of garden-variety fraud, its analysis would be in tension with both Yanakas and Plapinger. See, e.g., Yanakas, 7 F.3d at 317 (finding the disclaimer insufficient in part because it contained “no blanket disclaimer of the type found in Plapinger as to ‘any other circumstance which might otherwise constitute a defense’ to the Guarantee”). We agree with the District Court, however, that JPMorgan Chase is “an unusual and extreme case [that] ... provides little guidance.” 312 F.Supp.2d at 589, 286 F.Supp.2d at 358. Though characterized as a fraudulent inducement, the transaction in JPMorgan Chase bordered on a fraud in the factum— “the sort of fraud that procures a party’s signature to an instrument without knowledge of its true nature or contents.” Langley v. Fed. Deposit Ins. Corp., 484 U.S. 86, 93, 108 S.Ct. 396, 98 L.Ed.2d 340 (1987) (citing U.C.C. § 3-305(2)(c) cmt. 7 (1977)); see also Restatement (Second) of Contracts § 163 (1981) (referring to a fraud with respect to the “character or essential terms of a proposed contract”). The surety was asked to insure the delivery of a commodity, when in fact it was guarantying a loan. See JPMorgan Chase, 189 F.Supp.2d at 28. In such a case, where the party does not even know the “true nature” of what it is signing, it is unsurprising that the standards for effective waiver would be stricter, if waiver is possible at all. Cf. Restatement §§ 163—164 (explaining that a contract fraudulently induced is voidable, whereas a “contract” procured through fraud in the factum is no contract at all). Unlike Liberty Mutual, Royal does not seriously question the nature of the transactions covered by its policies. It characterizes SFC’s business as a sham, but it does not deny that SFC was issuing student loans and that the repayment of the interest and principal on those loans was insured by Royal’s policies. A fairly obvious risk of this insurance decision was that SFC would misrepresent the quality of the loans. For the reasons given above, the policies clearly assign this risk of fraud to Royal when they declare that its liability will be unaffected by a “breach of any representation” or “fraud with respect to the student loans.” Because the risk of fraud was so obvious, it is unimaginable that a party with Royal’s experience and knowledge would not have realized it was assuming that risk when it agreed to that language. Many of the remaining cases cited by Royal involved alleged violations of the federal securities laws. See, e.g., Caiola v. Citibank, N.A., 295 F.3d 312 (2d Cir.2002); *218 CP Kelco U.S., Inc. v. Pharmacia Corp., No. CIV.A.01-240, 2002 WL 31230816 (D.Del. Oct.2, 2002). It is well known that the federal securities laws provide broader fraud protection than the common law, having been enacted in response to the common law’s perceived failure at stamping out fraud in the securities markets. See, e.g., Herman & MacLean v. Huddleston, 459 U.S. 375, 388-89, 103 S.Ct. 683, 74 L.Ed.2d 548 (1983); In re Data Sys. Sec. Litig., 843 F.2d 1537, 1544 (3d Cir.1988). Since the analysis of cases like CP Kelco appears to have been colored by federal securities law, we do not believe those cases provide reliable guidance on this question of Delaware’s common law of contract. The lack of specificity in Royal’s waivers does not make them any less clear. As the Progressive Court put it, a “method of identification does not become unclear simply because it is terse.” 2002 WL 1558382, at *10. This is all the more true when the method of identification is hammered out by sophisticated parties aided by consummate legal professionals, who can be expected to anticipate the subjects it will identify. Accordingly, we predict that when sophisticated parties have inserted clear anti-reliance language in their negotiated agreement, and when that language, though broad, unambiguously covers the fraud that actually occurs, Delaware’s highest court will enforce it to bar a subsequent fraud claim. Royal asserts that this prediction risks giving protected parties a “license to commit fraud,” CP Kelco, 2002 WL 31230816, at *5, but the alternative poses its own danger. The danger is that a contracting party may accept additional compensation for a risk that it has no intention of actually bearing. This prevarication may amount to a fraud all its own. See Progressive Int’l, 2002 WL 1558382, at *10 (“In essence, Progressive is saying to the court, ‘Believe us now when we tell you we made a false promise to DuPont then.’ ”); Danann Realty Corp. v. Harris, 5 N.Y.2d 317, 184 N.Y.S.2d 599, 157 N.E.2d 597, 600 (1959) (“[T]he plaintiff made a representation in the contract that it was not relying on specific representations not embodied in the contract, while, it now asserts, it was in fact relying on such oral representations. Plaintiff admits then that it is guilty of deliberately misrepresenting to the seller its true intention.”). Given the potential for misrepresentation from each side of the agreement, the safer route is to leave parties that can protect themselves to their own devices, enforcing the agreement they actually fashion. This rule will make for less prolix disclaimers and reduce the likelihood that an intended allocation of the risk of fraud will be frustrated by an unintentional omission from a long and tedious list of representations. Such concerns have figured prominently in recent Delaware case law. See Kronenberg, 872 A.2d at 593 (accommodating “contractual freedom and efficiency concerns” as well as a “public policy ... intolerant of fraud”); Progressive Int’l, 2002 WL 1558382, at *10. When sophisticated parties include a broad but unambiguous anti-reliance clause in their agreement, the Delaware Supreme Court will likely indulge the assumption that they said what they meant and meant what they said. It follows from the foregoing discussion that the Royal waivers are in full force and effect. For the reasons given in Section A of this opinion, those waivers unambiguously strip Royal of the defense of fraud in the inducement. The District Court thus did not err in awarding the beneficiaries summary judgment on them, and it also did not err in denying Royal further discovery into the beneficiaries’ knowledge of *219 the fraud. The agreements unambiguously waive Royal’s fraud defense whether or not the beneficiaries had an inkling of the fraud before Royal, leaving Royal with no remaining defense to payment that discoverable evidence could support. To the extent set forth above, the judgment of the District Court is accordingly affirmed. C. We turn now to consider the scope of Royal’s obligations under the policies. The District Court ordered it to pay all claims submitted by Wells Fargo and Wilmington Trust, including claims that had been filed since the commencement of the lawsuit. Royal argues that the Court overlooked triable issues regarding whether all of the claimed losses were covered under the policies and whether the beneficiaries’ claims had been properly documented. The beneficiaries counter that they are entitled to immediate payment under the policies and that Royal’s defense based on SFC’s alleged misappropriation is waived along with its other fraud defenses. The scope of coverage is set forth in the first section of each policy. The language differs from policy to policy, but the substance is the same: Royal agrees to reimburse the insured or beneficiary under the policy for a “Loss” that occurs during the policy period. E.g., App. at 2247, 5479. Critically, “Loss” is defined as “for any Student Loan as to which a Claim is made in accordance herewith the Value as of the Default Date with respect to such Student Loan.” E.g., App. at 2017, 5527. The “Value” of a Student Loan is simply the “principal balance outstanding ... as of the Default Date plus accrued interest thereon.” E.g., App. at 2018, 5528. The “Default Date” is the first date of a “Default,” e.g., App. at 2016, 5526, which in turn is defined as: (i) a Student Loan becoming ninety (90) days or more delinquent (treating payments made by a Student but paid over to a bankruptcy court having jurisdiction over the Student as a preference item as not having been paid by the Student), with payments made after a delinquency applied to the earliest delinquency, or (ii) any impairment or avoidance of the rights of the Beneficiary or the Insured .in a Student Loan arising out of the bankruptcy or similar event or proceedings with respect to Student Finance Corporation or the Insured, including without limitation pursuant to Section 362 of the United States Bankruptcy Code. E.g., App. at 2032-33; see also, e.g., App. at 2016, 5479. These provisions give rise to a reasonable inference, if not an irresistible conclusion, that the “Loss” for which Royal agreed to reimburse the beneficiaries was the nonpayment of interest and principal on the student loans. If a loss may be traced to something other than a “Default” — that is, “a Student Loan becoming ninety (90) days or more delinquent” — it is not covered under the policies. 5 Under this highly reasonable interpretation, Royal need not reimburse the beneficiaries for nonreceipt of loan proceeds misappropriated by SFC. Wells Fargo argues that SFC’s misappropriation is just another form of fraud that Royal waived as a defense to payment. This argument misses the point. *220 Royal’s defense is not based on SFC’s fraud per se but on the extent of coverage afforded by the policies. The waivers themselves make this distinction clear. In the first eight policies, Royal agreed that its “obligation to pay any Claim” would be unaffected by a “breach of any representation.” In the last two policies, it agreed to waive any defense that would allow it “to avoid payment of its obligation under this policy.” Wells Fargo’s argument begs the question of whether a loss due to SFC’s conversion is one of Royal’s “obligation[s].” We also reject Wilmington Trust’s argument that its policies “require Royal to pay every Claim submitted to it upon the submission of a conforming ‘Notice,’ without more.” Wilmington Trust’s Br. at 40 (emphasis omitted). Although section III(A) of the Wilmington Trust policies provides that “[a]ny Claim ... shall be paid by the Insurer within sixty (60) days,” App. at 5481, 5528, other language appears to contemplate that Royal may challenge claims prior to payment. Section III(C)(1) requires the insured or beneficiary to deliver with notice of its claims a “written proof of loss form ... identifying the applicable Student Loans” and the unpaid balance on them. App. at 5481, 5528. The policies also require the insured to provide, upon Royal’s request, “evidence reasonably available with respect to circumstances surrounding a Loss.” Id. Requiring such evidence prior to payment would serve no purpose if Royal were obligated to pay any claim upon receipt. Arguably, section III(A) simply lays out a time frame for payment of a claim that the parties have already agreed is valid. To what extent SFC’s misappropriation contributed to the beneficiaries’ losses remains unclear. Royal points to evidence that Andrew Yao was receiving distributions from SFC at the rate of approximately $1 million per month while serving as its director, treasurer, and chief executive officer. See App. at 4405-07. According to Royal, SFC’s tottering financial position at the time makes it likely that some of this money was diverted from the income streams due the beneficiaries. The beneficiaries have not disputed this evidence on appeal, relying instead on the interpretive arguments rejected above. Having found the ten policies reasonably susceptible to an interpretation under which Royal would not be liable for losses due to SFC’s misappropriation, we conclude that further development of the record is necessary to determine the extent of this misappropriation before Royal may be ordered to pay the beneficiaries’ claims. On remand, the District Court must first decide whether or not the policies cover losses attributable to SFC’s alleged conversion. If not, it must then determine the extent to which SFC’s misappropriation contributed to those losses. Further discovery may be necessary, but we caution that this remand is not an opportunity for Royal to revive its rejected fraudulent inducement defense. The proceedings should be limited to deciding which of the losses claimed by the beneficiaries are covered under the policies. IV. After careful consideration of the parties’ arguments and submissions, we conclude that each of the ten Royal policies unambiguously and effectively waived its defense to payment of its obligations based on SFC’s fraud. The District Court correctly denied Royal rescission, correctly declared the policies in full force and effect, and correctly ordered Royal to perform its obligations under them. That part of the District Court’s decision is affirmed. *221 The remainder of the decision is vacated. Triable issues regarding the scope of the policies’ coverage and the nature of the beneficiaries’ losses preclude summary-judgment. On remand, the District Court must determine whether the policies cover all of the losses claimed by the beneficiaries before ordering Royal to pay them. Footnotes [1] A third appellee-beneficiary, PNC Bank, N.A., reached a settlement with Royal prior to our decision. This opinion does not discuss its policy with Royal or its claims thereon. [2] MBIA issued a "Certificate Guaranty Insurance Policy” guarantying the student loans held by the Wells Fargo trusts. Royal and MBIA disagree about the function of this guaranty. MBIA characterizes it as backstop insurance for a risk primarily insured by Royal, while Royal characterizes it as supplemental coverage protecting the trusts from risks not covered under the Royal policies. Royal no longer disputes, however, that MBIA has contractual standing to sue, so we assume that it is a proper plaintiff-appellee. [3] In Policy RST 147538, clause (C) appears as clause (D) and follows a new clause (C), which leaves Royal liable irrespective of "ANY FAILURE ON THE PART OF THE INSURED, THE SERVICER OR THE BENEFICIARY TO OBSERVE OR PERFORM ANY COVENANT OR CONDITION.” App. at 2254. [4] Royal asserts, citing several Third Circuit cases, that this Court always considers extrinsic evidence in determining whether an agreement is unambiguous. These cases, however, were not decided under Delaware law. See, e.g., In re New Valley Corp., 89 F.3d 143, 149 (3d Cir.1996) ("[a]pplying the federal common law of contract”); Mellon Bank, N.A. v. Aetna Bus. Credit, Inc., 619 F.2d 1001, 1005 (3d Cir.1980) (Pennsylvania law). Although no text can be read in a vacuum, the Delaware Supreme Court has held that in determining whether an ambiguity exists a court may consider only "undisputed background facts to place the contractual provision in its historical setting.” Eagle Indus., Inc. v. DeVilbiss Health Care, Inc., 702 A.2d 1228, 1232 n. 7 (Del.1997). "[U]ndisputed background Even if we considered the sworn affidavit of Royal's employee, William Hibberd, it would not lend ambiguity to the plain language of the Royal policies. Hibberd could state only that it was his “understanding” that the policy language was not intended to leave Royal liable in the case of a "widespread fraud like the one that occurred here.” E.g., App. at 5345. Hibberd's subjective impression of what the parties meant, recollected years after the events in question, is at very best a "scintilla” of evidence in support of the objective reasonableness of Royal's alternative construction of the policies. [5] Wé recognize that by their terms the policies also extend coverage to losses resulting from the impairment of the beneficiaries' rights in bankruptcy. SFC entered bankruptcy in June 2002, but the parties have not discussed how this event affects their rights and obligations under the policies. We leave this interesting question to the District Court.